Scribd Logo
Upload

Welcome to Scribd!

  • 0 views

    Stock Prediction Enhanced Presentation

    Uploaded by

    Rishi Chourasia

    Copyright:

    © All Rights Reserved
    Download as PPTX, PDF, TXT or read online from Scribd

    Stock Prediction Enhanced Presentation

    Uploaded by

    Rishi Chourasia
    0 views12 pages

    Copyright

    © © All Rights Reserved

    Available Formats

    PPTX, PDF, TXT or read online from Scribd

    Did you find this document useful?

    Is this content inappropriate?

    Copyright:

    © All Rights Reserved
    Download as PPTX, PDF, TXT or read online from Scribd
    Download as pptx, pdf, or txt
    0 views12 pages

    Stock Prediction Enhanced Presentation

    Uploaded by

    Rishi Chourasia

    Copyright:

    © All Rights Reserved
    Download as PPTX, PDF, TXT or read online from Scribd
    Download as pptx, pdf, or txt
    of 12

    Stock PriceTraining

    Prediction
    Seminar
    using Report
    Time Series and NLP
    Rishi Chourasia, IIT Roorkee
    Abstract
    his study addresses the challenge of stock price prediction by using ARIMA and LSTM models. The
    esearch focuses on comparing the accuracy of different forecasting models using the SPY ETF dataset
    ghlighting the superior performance of the LSTM model.
    Introduction
    k price prediction is influenced by political, economic, and financial factors. This research focuses
    pplying advanced machine learning models such as LSTM for stock price prediction using historical d
    Research Methodology
    Data Collection: The study utilizes the SPY ETF dataset, tracking the S&P 500 Index with various data
    points (opening, closing, high, low prices, trading volumes).
    Data Preprocessing
    Steps:
    - Data cleaning: Missing values and outliers handled.
    - Feature engineering: Creating moving averages, volatility indices.
    - Train-Test split: 70-80% for training, 20-30% for testing.
    Models Used
    1. Naive Forecast Model
    2. Moving Average Models (SMA, EMA)
    3. ARIMA Model
    4. Neural Networks (Linear, Dense, RNN, LSTM)
    5. Hybrid CNN-RNN Models
    Results: Model Comparison
    STM model demonstrated the lowest Mean Absolute Error (MAE), followed by ARIMA and simple m
    aive Forecast and Moving Average.
    Conclusion
    dy demonstrates that LSTM models significantly improve prediction accuracy compared to traditiona
    like ARIMA and moving averages. Future research could explore hybrid models combining CNNs and
    Future Work
    ploration of hybrid models combining CNN for feature extraction and LSTM for time-series forecasti
    pplication of advanced neural network architectures for further improvement in prediction accuracy
    Abstract
    research aims to predict stock prices using models like ARIMA and LSTM, emphasizing LSTM's superior
    Introduction
    ock market's non-linear behavior necessitates advanced predictive techniques for accurate stock price f
    Methodology
    a from SPY ETF was preprocessed, including handling missing values and creating features like moving a

    You might also like